The sonic signature of jitter and how to conquer it Part II: anti-jitter tweaks

art · « **Reply #40 on:** 19 Mar 2008, 01:50 pm »

A lot of stuff to address........

Vibrations and jitter:

I have measured the effects of crystal sensitivity to vibrations. Not a big a problem as you might think. Does not mean that reducing vibrations does not have other advantages. Just doesn't do much to jitter. Something else is going on.

The power supply:

If you read the link on oscillator phase noise, you will run across a term called FM flicker noise. The Q of the crystal, while high, is not infinite. Noise, including PSU garbage, can FM modulate the signal. It has to be clean, or else you get sidebands.

DVD players have nasty clock schemes that involve fractional-N PLLs and other stuff to come up with all the clock frequencies required. The problem in one of them is worse than in CD player. A lot more stuff going on inside........switching supplies........PLL clock gens.....

Not saying it won't help. I am saying it is not that big of a problem in stand-alone CD players.

SPDIF:

It is flawed. Period.

You will have 2 things that enter into jitter. One is the inherent jitter, related to the PLL. That is a number that will always be there. Can't go lower without some "fix" downstream. The second is additional jitter that is dumped on top of that by impedance mismatches.

Cables not only affect the amount of the mismatch, but depending on the length, it could ameliorate it by making the reflections arrive after the decision point.

The "best" way to deal with jitter must first start with the cause. The simplest way is a stand-alone CD player. Less issues to deal with.

SPDIF......lots of ways. Outboard devices that have been discussed can solve most of the problems. The only real way is to buy DACs that are designed to fully deal with the issue. It is not something simple that the average listener can do themselves.

Pat

*Scotty* · « **Reply #41 on:** 19 Mar 2008, 03:53 pm »

Pat, can you short list the products that you feel have been designed in such a way as to ameliorate the problems due to the SPDIF interface.
Scotty

art · « **Reply #42 on:** 19 Mar 2008, 06:30 pm »

I have absolutely no idea who makes what, or why. I would say that nothing under 2 metres has much of a chance.

I know that the cable we make does. There is an explanation on that somewhere on our site. (Temporarily out of parts to make more right now.)

Pat

Geardaddy · « **Reply #43 on:** 19 Mar 2008, 07:59 pm »

Recent Audiogon verbage regarding some new wave USB DAC:

"the USB DAC that I have been citing does NOT connect to a CD transport, It connects directly to a computer i.e. my posts statements regarding the "hard drive" . With this methodology, the computer does not have a SPDIF (traditional) connection whatsoever. It streams pure data directly from the hard drive, WITHOUT the subsequent clock data. It is much different with this methodology. (It is also much more elegant). This is the ONLY way the sonic improvements that I am discussing can have any meaning.....the complete elimination of having to carry clocking information from the CD transport to the DAC is where the huge jump factor relies (remember, the music data is now on a hard drive). With this in mind, any simplification of transferring the music data from point A to point B really does yield sonic wonders."

A good way to think of exotic reclocking devices (such as the Esoteric) is they are repairing the myriads of synchronization issues associated with the abysmal SPDIF connection. With a SPDIF's flawed approach, the clocking information is interleaved with the music data. The Esoteric Rubidium's incredible improvements seem to prove how poor the idea of interleaving clock information along with the music data has been all along! Of course, we can do this much more elegantly if we don't even try to interleave the clocking data on top of music in the first place. Can we say "USB"

What do you guys think of this?

opaqueice · « **Reply #44 on:** 20 Mar 2008, 01:01 am »

Quote from: Geardaddy on 19 Mar 2008, 07:59 pm

USB is one way to go, although my impression is that it's very hard to do properly. Another option, as I mentioned before, is a squeezebox, transporter, etc.

Those devices receive data over a computer TCP/IP network - which is asynchronous and hence has no jitter - and then DAC it using a local oscillator as a clock. S/PDIF never enters the equation unless you connect to an external DAC.

Geardaddy · « **Reply #45 on:** 20 Mar 2008, 01:59 am »

Agreed. The guys were talking about a new USB technology as if it represents a jitter-free a paradigm shift and its not....I believe what they are referring to is an asynchronous DAC such as Wavelength....

TomS · « **Reply #46 on:** 20 Mar 2008, 02:09 am »

Not sure what you mean by "new", but perhaps they're talking about these from Gordon Rankin (Wavelength)?

http://www.usbdacs.com/Concept/Concept.html

jkeny · « **Reply #47 on:** 20 Mar 2008, 02:14 am »

There is a new 8 channel Dac from ESS called a Sabre which claims to have eliminated jitter in a new form of PLL - I know about manufacturer claims but this is posted by the design engineer responsible for the DAc & samples seem to stand up to the claims! http://www.diyaudio.com/forums/showthread.php?s=&threadid=117238

So maybe an end to SPDIF jitter?

Geardaddy · « **Reply #48 on:** 20 Mar 2008, 02:41 am »

Quote from: TomS on 20 Mar 2008, 02:09 am

Not sure what you mean by "new", but perhaps they're talking about these from Gordon Rankin (Wavelength)?

http://www.usbdacs.com/Concept/Concept.html

Tom, apparently its not Wavelength line...this portion of the discussion starts towards the end of the thread...

"As far as the "Beta" pre-production USB DAC that I heard though (from a company other than Gordan Rankin's) it was a serious jaw dropping experience!"

http://forum.audiogon.com/cgi-bin/fr.pl?ddgtl&1203194310&openfrom&32&4#32

Quote from: jkeny on 20 Mar 2008, 02:14 am

There is a new 8 channel Dac from ESS called a Sabre which claims to have eliminated jitter in a new form of PLL - I know about manufacturer claims but this is posted by the design engineer responsible for the DAc & samples seem to stand up to the claims! http://www.diyaudio.com/forums/showthread.php?s=&threadid=117238

So maybe an end to SPDIF jitter?

Hmmm...interesting. Don't know. Ehider is appropriately tight lipped about it....

Geardaddy · « **Reply #49 on:** 20 Mar 2008, 04:14 am »

For those interested, here are the jitter wav files that Ethan Winer promised us:

http://www.audiocircle.com/index.php?topic=52363.new#new

AphileEarlyAdopter · « **Reply #50 on:** 20 Mar 2008, 11:08 pm »

Geardaddy, see opaqueice's post. That is the best but simple explanation.
The guy hyping the new usbdac in audiogon, is just impressed with the sound. It is not architecturally a new thing - the async dacs from Wavelength and the Squeezebox/transport have all explored this territory before. If you ask me, the Squeezebox + Empirical Audio Pace car might be the best transport that uses a computer. We really dont know how the multi-thousand dollar transports compare against the Squeezebox+Pace car. It is quite possible they are probably equal in jitter performance.

Geardaddy · « **Reply #51 on:** 21 Mar 2008, 01:14 am »

EarlyAudioAdapter, I understand and agree with opaqueice's comments. I also agree with you in regards to the Pacecar concept. I know Steve Nugent from Empirical Audio is coming out with his F1 DAC in a year or so, which is a pacecar and dac in one package. A tubed version will follow after that. Anyway, I forwarded Steve the thread about this "new" USB technology and here's what he had to say:

"USB indeed does transfer a clock, similar to the way that S/PDIF has an embedded clock. The primary difference between the two is that the chips available for clock recovery for USB are actually better at jitter rejection than the receiver chips for S/PDIF clock recovery. That is the primary advantage of USB, as well as an apparent lower sensitivity over long USB cables compared to long S/PDIF cables.

And the Synchronization comment is false. These interfaces all use forms of Phase-locked-loops or delay-locked-loops. The rubidium clock or a Superclock are just providing a low-jitter clock for the PLL. Nothing to do with "synchronization". Both USB and S/PDIF have embedded clocks.

An I2S interface actually has separated clocks, so this is a big advantage and results in even lower jitter than USB or S/PDIF."

AphileEarlyAdopter · « **Reply #52 on:** 23 Mar 2008, 06:26 pm »

Ofcourse, Steve is right. The ASYNC USB is a new development. I knew USB had some issues (before ASYNC USB) that is why I went the Squeezebox. But I had to get it modded (the first time I am going for mods) because the stock one's jitter was not good enough for me. The Squeezebox is really a async sound card. It is not dependent on the computer power supply etc. But you are limited by the accuracy/jitter of the Squeezebox itself. No wonder high performance costs lots of money (even though I think it should not cost like $10K or so).

(I thought I was done with the modded power supply for the Squeezebox, but now it seems the DV 5V input is converted to various other voltages inside, so now there is a mod to supply really clean voltages for all these from outside..it costs another $300 or $400 oh man ..this will never end ..)

Geardaddy · « **Reply #53 on:** 23 Mar 2008, 06:31 pm »

AphileEarlyAdopter, I will conducting a little A+B, blinded, listening research in the not too distant future. The participants are still TBD, but I plan in my mind on the following:

1. Granite Audio 657 with tube output stage and no pre-amp...very sweet
2. modded SB3 (Bolder)
3. modded sonos + pacecar + dac (empirical audio)
4. sonos with USB conversion adapter + USB DAC from Exemplar...

Did you ever hear the Slim Devices TP +/- modwright mod before landing on the SB3?

AphileEarlyAdopter · « **Reply #54 on:** 24 Mar 2008, 10:50 am »

Geardaddy, you have some cool gear and you probably chose that moniker appropriately yourself

While experts like Robert Harley, have tried to defined the sonic signature of jitter. I would characterize the sound as 'tending towards the sound of LP'. Even with my entry level MMF-5 my analog setup has so much 'soundstage' or more accurately 'recording venue ambience'. This aspect of the recording seems pretty difficult to recreate with digital. ie. when you hear it played you can hear the individual instruments separately and the space around them. It is not one mass of 'gumbo'. A simple rule of thumb, different CDs sound very VERY different.
I have an idea, of the sources you have chosen, when it comes to neutrality which one will win. But I do not want to bias you

Good luck with you A/B. I envy the fun you are gonna have.
Please report in detail, I will take some vicarious pleasure in reading it.
(If anybody in San Francisco Bay Are doing these kind of tests, I would like to participate.)

Geardaddy · « **Reply #55 on:** 27 Mar 2008, 02:07 pm »

Hey there AphileEarlyAdopter. Actually I have sts9fan http://www.audiocircle.com/index.php?topic=50144.80 to thank. It was his stroke of genius in the midst of that "speaker showcase" thread that just had me howling with laughter. It was so good, I just had to memorialize it. I believe its a freezeframe from a Jeff Bridge's movie whose name escapes me at present. As far as your comments about digital:

Quote from: AphileEarlyAdopter on 24 Mar 2008, 10:50 am

"Even with my entry level MMF-5 my analog setup has so much 'soundstage' or more accurately 'recording venue ambience'. This aspect of the recording seems pretty difficult to recreate with digital. ie. when you hear it played you can hear the individual instruments separately and the space around them. It is not one mass of 'gumbo'."

I concur 100%. Your description is more poetic than Harley's, but it pinpointed exactly what "I/we" think jitter is doing. It is the recreation of the recording venue with its ambient, 3D, spatial cues, etc that I crave (along with of course the warm, organic, liquid midrange that only tubes and vinyl appears to create....). My Granite Audio CDP gets very, very close to pulling this stuff off and that is why I have been relunctant to convert quite yet to computer audio. I am waiting to really here "this"....

gitarretyp · « **Reply #56 on:** 27 Mar 2008, 02:22 pm »

Geardaddy, the avatar is from The Big Lebowski

Geardaddy · « **Reply #57 on:** 27 Mar 2008, 03:53 pm »

oh, okay....that is what I thought but was not sure....

audioengr · « **Reply #58 on:** 4 Apr 2008, 05:34 pm »

Quote from: BrianM on 18 Mar 2008, 03:06 pm

Quote from: art on 18 Mar 2008, 03:02 pm
Only works for SPDIF. Which is the only application that jitter needs to be addressed. (I already showed in the other thread that jitter in a stand-alone player is not much of a problem. Stop worrying about it unless you have lots of time and money. Mostly the latter.)

I probably missed it somewhere, but to what extent do you feel the SPDIF problem is ameliorated by a so-called optimized digital cable length?

It certainly helps. Read this white-paper:
http://www.positive-feedback.com/Issue14/spdif.htm

audioengr · « **Reply #59 on:** 4 Apr 2008, 05:37 pm »

Quote from: BrianM on 19 Mar 2008, 10:31 am

Quote from: *Scotty* on 19 Mar 2008, 01:17 am
SPDIF cannot be fixed or cured because it is inherently flawed.

That may well be, but like a lot of dogmatic statements it might just suffer from a narrow perspective. The transducers in your speakers can't be "fixed" or "cured" either, and are inherently flawed. (So is your listening room.) Are we losing more sleep over these facts, or is it more about the jitter? Which is worse, an awesome DAC + SPDIF or an awesome CDP which is probably compromised in other ways? I'm sure nobody's against a better solution than SPDIF being more widely accepted, but the question (I thought) is whether SPDIF can be ameliorated to below the point at which other factors become more relevant to improving the sound. Or, perhaps, whether it is already below such concerns.

I was of the opinion that S/PDIF was a lost cause and as a result did a lot of I2S interfaces. With the advent of my Pace-Car reclocker however, I have discovered that S/PDIF can approach the jitter of I2S assuming that the interfaces are well-implemented and properly terminated etc...

Steve N.